Magistral Medium 1.2 — benchmark results

Provider: Mistral. Released 2025-06-10. Access: API.

Unified ELO 1593 ± 15, rank #547 of 1839 rated models, from 76 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA LiveCodeBench75.03Pass@1 (%)86.5
AA AIME 202582Accuracy (%)77.5
AA MMLU-Pro81.51Accuracy (%)75.4
Wolfram LLM Benchmarking Project51.7Correct Functionality (%)72.8
Epoch AI - Scicode39.24Score68.8
AA Omniscience-26.23Score65.4
AA Omniscience - Health22.9Accuracy (%)65
AA Global-MMLU-Lite - Bengali83.42Accuracy (%)61.9
AA Omniscience - Science, Engineering & Mathematics29.5Accuracy (%)61.7
AA Omniscience - Software Engineering (SWE) - Go22Accuracy (%)61.2
AA Global-MMLU-Lite - Arabic84.83Accuracy (%)61
AA Global-MMLU-Lite - Hindi83.75Accuracy (%)60.3

Interactive version: theaggregate.ai/model?slug=magistral-medium-1-2 · How It Works · Data refreshed daily, snapshot 2026-08-05.